Aberrant eukaryotic translation initiation factor 4E-dependent mRNA transport impedes hematopoietic differentiation and contributes to leukemogenesis

Abstract

The eukaryotic translation initiation factor 4E (eIF4E) acts as both a key translation factor and as a promoter of nucleocytoplasmic transport of specific transcripts. Traditionally, its transformation capacity in vivo is attributed to its role in translation initiation in the cytoplasm. Here, we demonstrate that elevated eIF4E impedes granulocytic and monocytic differentiation. Our subsequent mutagenesis studies indicate that this block is a result of dysregulated eIF4E-dependent mRNA transport. These studies indicate that the RNA transport function of eIF4E could contribute to leukemogenesis. We extended our studies to provide the first evidence that the nuclear transport function of eIF4E contributes to human malignancy, specifically in a subset of acute and chronic myelogenous leukemia patients. We observe an increase in eIF4E-dependent cyclin D1 mRNA transport and a concomitant increase in cyclin D1 protein levels. The aberrant nuclear function of eIF4E is due to abnormally large eIF4E bodies and the loss of regulation by the proline-rich homeodomain PRH. We developed a novel tool to modulate this transport activity. The introduction of IkappaB, the repressor of NF-kappaB, leads to suppression of eIF4E, elevation of PRH, reorganization of eIF4E nuclear bodies, and subsequent downregulation of eIF4E-dependent mRNA transport. Thus, our findings indicate that this nuclear function of eIF4E can contribute to leukemogenesis by promoting growth and by impeding differentiation.

FIG. 1.

Levels, subcellular distribution, and interaction of eIF4E and PRH are altered in M4 AML, M5 AML, and bcCML specimens. (A) Western blot analysis of whole-cell extracts in cells derived from specimens as indicated. β-Actin is shown as a control for protein loading. (B) Whole-cell lysates were immunoprecipitated with anti-eIF4E Ab (IP eIF4E), and the resulting Western blot was probed for PRH. IP, immunoprecipitated fraction; s, supernatant after immunoprecipitation; W.B., Western blot. (C) In the left panel are shown confocal micrographs of cells stained with FITC-conjugated anti-eIF4E Ab (shown in green); anti-PRH Ab, followed by Texas red-conjugated anti-rabbit IgG Ab (shown in red); and anti-PML Ab (5E10), followed by Cy5-conjugated anti-mouse IgG Ab (shown in blue). The PML-eIF4E overlay is shown in light blue, the PML-PRH overlay is shown in pink, the PRH-eIF4E overlay is shown in yellow, and the triple eIF4E-PML-PRH overlay is shown in white. The objective was 100x with a further magnification of 2 (A-H) or 3 (I-X) fold. In the right panel, cells were fractionated into cytoplasmic and nuclear compartments and analyzed as indicated. β-Actin and SC 35 were used as a loading control for the cytoplasmic and nuclear fractions, respectively. BM, cells derived from the healthy individuals. Other specimens are as indicated. AML-ETO indicates that translocation was found in that M2 specimen (Table 1).

FIG. 2.

Cyclin D1 levels are posttranscriptionally upregulated in M4 AML, M5 AML, and bcCML patients. (A) Whole-cell lysates were analyzed by Western blot (W.B.) as indicated. The levels of cyclin D1 are increased in all leukemia specimens (lanes 2, 3, and 5 to 8). β-Actin is shown as a control for protein loading. (B) Northern blot (N.B.) analysis of whole-cell lysates shows upregulation of cyclin D1 mRNA levels in cells derived from M1 AML, M2 AML, M2/AML-ETO AML, and ALL patients (lanes 4 to 7). Cyclin D1 mRNA levels were not altered in M4 AML, M5 AML, and bcCML specimens (lanes 2, 3, 9, and 10). (C) RNA, isolated from nuclear (n) and cytoplasmic (c) fractions, was analyzed by Northern blot as indicated. Nuclear export of cyclin D1 mRNA was increased in the cells derived from M5 AML (lanes 3, 4, 9, and 10) and bcCML (lanes 5 and 6) patients. tRNA^Lys and U6snRNA were used as markers for the cytoplasmic and nuclear fractions, respectively. GAPDH is shown as a control for RNA loading. Samples are labeled as described in Fig. 1.

FIG. 3.

Expression of IκB-SR in CD34⁺ cells, derived from M5 AML and bcCML patients, correlates with the downregulation of the c-myc expression and the restoration of the expression and subcellular distribution of eIF4E and PRH proteins. (A) Cells were stained with anti-eIF4E Ab, followed by Texas red-conjugated anti-mouse IgG Ab (shown in green) and anti-PRH Ab, followed by Cy5-conjugated anti-rabbit IgG Ab (red). Nuclei were stained with DAPI (blue in panels A and E; gray in panels I and M). The PRH-eIF4E overlay (ov.) is shown in yellow. The objective was a 100× lens with a further magnification of two (A to L)- or three (G to R)-fold. (B) Western (W.B.) and Northern (N.B.) blot analysis of CD34⁺ cells derived from the healthy individuals (BM) and M5 AML patients [AML(M5)-IκB]. Blots were probed as indicated. (C) IκB-SR expression correlates with the restoration of cyclin D1 mRNA transport. Northern blot analysis of RNA isolated from nuclear (n) and cytoplasmic (c) fractions of CD34⁺ cells derived from healthy individuals (BM) and M5 AML patients. In contrast to Ad-GFP-transduced cells (−IκB-SR), CD34⁺ M5 AML cells that express IκB-SR (+IκB-SR) showed the same subcellular distribution of cyclin D1 mRNA (lanes 5 and 6) as CD34⁺ BM cells. tRNA^Lys and U6snRNA were used as markers for cytoplasmic and nuclear fractions, respectively. −IκB-SR, cells transduced with Ad-GFP; +IκB-SR, cells transduced with Ad-GFP-IκB-SR. (D) Western blot analysis reveals upregulated levels of c-myc in M4 AML, M5 AML, and bcCML specimens (c-myc W.B.). β-Actin is shown as a control for protein loading.

FIG. 4.

eIF4E blocks differentiation mediated by ATRA and vitamin D₃. (A) Vector controls or cells expressing wild-type eIF4E, W56A mutant eIF4E, or W73A mutant eIF4E, as indicated, were treated for 5 days with either a vehicle control (untreated, white histogram), ATRA (red histogram), or 1,25(OH)₂D₃ (blue histogram) and then analyzed for cell surface marker expression. The median fluorescence intensity for each histogram (m) is given in each panel. The results are representative of three independent experiments. (B) Cells from panel A were stained with Wright-Giemsa, and changes in morphology were evaluated. Untreated cells shown are the vehicle control for vitamin D₃; no effect on differentiation was seen with either carrier.